Precision in motion control and perception is critical to survival - whether prey running over rough terrain or a pilot landing on an aircraft carrier. Behavioral precision depends on precision at many levels - e.g. sensory transduction, motor units, and the central nervous system - and there are open questions about which level(s) limit overall precision. As fundamental examples, we don't know whether the majority of vestibuloocular reflex (VOR) imprecision originates in the periphery, the CNS, or the oculomotor system. Similarly, we do not know whether the majority of perceptual imprecision originates in the periphery or CNS. The long-term general goal of this research is to understand the role of precision in sensation, behavior, pathology, clinical diagnosis and neural processing. The relative simplicity of the vestibular system makes it an excellent model to study the basic principles of how the brain processes noise, which may be more broadly applicable to other sensory systems. The short-term goal of this work is to understand precision in vestibular sensorimotor reflexes and perception using novel techniques to measure and isolate sources of imprecision. For example, the sensitivity of clinical tests at diagnosing particular disorders coul be strengthened if sources of imprecision could be isolated. To achieve this goal, we propose the following specific aims. Aim 1: Study sensory, motor and perceptual noise by simultaneously assaying VOR and perceptual precision in humans. To determine the extent to which the VOR and perceptual thresholds depend on a shared noise source, we propose to measure perceptual and VOR responses simultaneously using threshold-level stimuli, and quantify trial-by-trial co-variation. We hypothesize that a shared, presumably peripheral sensory noise source underlies both perceptual and oculomotor responses for yaw rotation and inter-aural translation. Aim 2: We will separate sensory and motor precision by exploiting a unique characteristic of the translational VOR - that the sensitivity of the response can be modulated by fixation distance - allowing the effective decoupling of the sensory and motor components of the pathway. We will measure trial-to-trial VOR variability in response to repeated stimuli, while presuming that VOR variability dependent on sensory imprecision will scale with VOR sensitivity. We hypothesize that sensory noise dominates translational VOR variability.